customer-segmentation-word-concepts-banner-vector.jpg

Customer Personality Analysis

Customer Personality Analysis is a detailed analysis of a company’s ideal customers. It helps a business to better understand its customers and makes it easier for them to modify products according to the specific needs, behaviors and concerns of different types of customers.

Customer personality analysis helps a business to modify its product based on its target customers from different types of customer segments. For example, instead of spending money to market a new product to every customer in the company’s database, a company can analyze which customer segment is most likely to buy the product and then market the product only on that particular segment.

FEATURES

People

Products

Promotion

Place

Target

Dataset Resource: [https://www.kaggle.com/imakash3011/customer-personality-analysis]

0.CONTENT

1. Exploratory Data Analysis

1. Data Preview and Variable Corrections

2. Distribution of Variables

1. Distribution of Numeric Variables 2. Distribution of Categorical Variables

3. Crossover of Variables

1. Crossover of Categorical Variables 2. Crossover of Numeric Variables

2. Clustering to Summarize Customer Segments.

Data Preprocessing

1. Data Cleaning

1. Missing Data
2. Outlier and Noisy Data

2. Data Standardization

3. Feature Selection

4. Clustering

5. Conclusion

1. Exploratory Data Analysis

1.Data Preview and Variable Corrections

2. Distribution of Variables

1. Distribution of Numeric Variables

Inferences

2- Distribution of Categorical Variables

3. Crossover of Variables

1- Crossover of Categorical Variables

Now we will briefly interpret the tables of the variables that are important to the company one by one;

MntWines: There seems to be an excess of master's education level living alone in wine consumption. The use of wine is also high among those who are university graduates and live as "Absurd" and "Divorced". However, wine use is remarkably low among University graduates and PhD graduates living alone. When we look at the table in general, the use of wine increases as the level of education increases. However, we do not see any interesting variation in wine consumption according to marital status.

However, the number of "Absurd" customers of the company that we should pay attention to here is very small. The company should take its steps with this low level in mind.

MntFruits: If we look at the fruit consumption of the company customers, I do not think that it varies according to the education level. Likewise, marital status does not affect much. However, the number of university graduates who are "Absurd" is still high. Then there is the increase in fruit consumption among customers at the "Widow" and "2nCycle" education level. Customers living alone have low fruit purchasing levels.

MntMeatProducts: The level of meat seems to increase according to the education level of widowed and single customers here. It has progressed in the opposite direction to other lifestyles. Meat purchasing levels of customers living alone are low.

MntFishProducts: In this table, we can see that people with "2n Cycle" education level who are widowed and living together have a high fish consumption. Other rods are close to each other, so there are no differences in fish consumption according to education level and lifestyle. Again, the rate of purchasing fish for those who live alone is low, and those of "Absurds" are high.

MntSweetProducts: We see that widows and people with "2nCycle" education level are high in sweet purchases. What draws our attention here is that as the level of education increases, the consumption of sweets decreases at the same level. But we can also see a significant difference between widows and people with other lifestyles.

MntGoldProds: We cannot say that gold purchase has anything to do with education level or marital status. Yes, there is an increase in the Master level in widows, but I do not think it is at a level that requires serious planning for the company. Customers with an "absurd" lifestyle are again at a high level.

NumDealsPurchases: According to the marital status of the customers shopping with a discount, we can see the differences in usage at the "Basic" education level. In fact, most of the variables do not seem to intersect, except for the University Graduate and Master education levels. This shows that there are many different discount usage preferences between education level and lifestyles in the use of the number of discounts.

NumWebPurchases: In this table, we can clearly see that as the level of education increases, the number of customers shopping on the company's website increases. There is not much difference between marital status (except for people living alone)

NumCatalogPurchases: There is no difference in the use of catalogs according to education level. The same can be said for lifestyles. There is not much difference other than those who live alone and those who are in the "Absurd" state. Those with only "Basic" education level are less likely to shop through the catalog.

NumStorePurchases: Interestingly, as the level of education increases, the number of purchases made from the store also increases. As a lifestyle, we can say that widows shop more than the store. Those who live alone, their store shopping decreases at the "phD" education level.

NumWebVisitsMonth: The education level of the visitors of the company's website that caught our attention is at "Basic" level. In the previous table, the level of shoppers was low at "Basic" education levels. There is not much change in other educational backgrounds and lifestyles. Individuals living alone have a slightly higher level of website visits, and university graduates have a lower level of website visits than others.

AcceptedCmp3: We see that people living together and alone prefer Campaign 3 more. However, there is no difference in the use of campaigns according to educational status.

AcceptedCmp4: If we look at Campaign4, the usage of campaign 4 changes according to the lifestyles of people in the "2n Cycle" education status. Campaign 4 usage of people in the "Divorced" and "Widow" lifestyle is high. The use of those who live alone is low. There does not appear to be much variation by education level.

AcceptedCmp5: There does not appear to be a change in the use of campaign 5 according to education level and lifestyle. Except for "absurd" lifestyles and widows.

AcceptedCmp1: We see that widows are high in campaign1 usage, there is no difference in education level. Except for "absurd" lifestyles. But it should not be forgotten that the company has very few customers with the "Absurd" lifestyle.

AcceptedCmp2: Customers with "Divorced" life style and "2nCycle" training have high campaign2 usage. Likewise, the use of campaign 2 increases in almost every lifestyle at the Phd education level.

Complete: In the last 2 years, the level of customers complaining has increased among customers with "2nCycle" education level, but it generally decreases as the education level increases.

Last_Campaign: The rate of use of the last campaign generally increases as the level of education increases, and at the same time, the rate of use by widows is high. The use of people with phD education level living alone is also significantly high. Usage rates vary considerably according to lifestyles.

2- Crossover of Numeric Variables

2. Clustering to Summarize Customer Segments

1. Data Preprocessing

1- Data Cleaning

1- Missing Data

Thus, we can get maximum clustering values by assigning the most realistic values to "NaN" values.

Outlier and Noisy Data

2. Data Standardization

3. Feature Selection

PCA(Principal Component Analysis)

4. Custering

5. Conclusion